The maize gene liguleless2 encodes a basic leucine...

The maize gene liguleless2 encodesa basic leucine zipper protein involvedin the establishment of the leafblade–sheath boundaryJustine Walsh, Cynthia A. Waters, and Michael Freeling1

Department of Plant and Microbial Biology, University of California, Berkeley, California 94720 USA

The blade and sheath of a maize leaf are separated by a linear epidermal fringe, the ligule, and two wedge-likestructures, the auricles. In plants homozygous for the null mutation, liguleless2-reference (lg2-R), the liguleand auricles are often absent or positioned incorrectly and the blade–sheath boundary is diffuse. Thisphenotype is in contrast to that of liguleless1-reference (lg1-R) mutant plants, which have a more definedboundary even in the absence of the ligule and auricles. Additionally, mosaic analysis indicates the lg2-Rphenotype is cell-nonautonomous and the lg1-R phenotype is cell-autonomous. Using scanning electronmicroscopy we show that lg2-R mutant plants are affected before the first visible sign of ligule and auricleformation. We have cloned the Lg2+ gene through a Mutator8 transposon insertion allele, and verified it withfive independently derived alleles. The comparison of genomic DNA and cDNA sequences reveals an openreading frame encoding a protein of 531 amino acids with partial homology to a subclass of plant basic leucinezipper (bZIP) transcription factors. Although a large body of molecular and biochemical characterization existson this subclass of bZIP proteins, our work represents the first report of a mutant phenotype within thisgroup. A specific reverse transcriptase (RT)–PCR assay shows LG2 mRNA expression in meristem/developingligule regions. RT–PCR also shows that LG2 mRNA accumulation precedes that of LG1 mRNA. The mutantphenotype and expression analysis of lg2 suggest an early role in initiating an exact blade–sheath boundarywithin the young leaf primordia.

[Key Words: Maize; plant; bZIP; liguleless2; leaf development; ligule]

Received September 23, 1997; revised version accepted November 6, 1997.

The maize (Zea mays) leaf contains three regions. Themajority of the mature leaf consists of the basal sheathand the distal blade. Positioned between these two re-gions is the ligular region, represented by a linear epider-mal fringe, the ligule, and two wedge-like structures, theauricles. All three regions contain epidermal, ground,and vascular tissues that are continuous with each otherbut distinct in cell types and patterns (Sharman 1942;Sylvester et al. 1990). The ligular region therefore repre-sents a boundary within the developing leaf primordium.

The developmental stages of the ligular region havebeen well characterized. The morphology and divisionpattern of ligular region epidermal cells were determinedhistologically (Sharman 1942) and by scanning electronmicroscopy (SEM) analysis (Becraft et al. 1990; Sylvesteret al. 1990). Initially, cells in the position where the ligu-lar region will develop divide anticlinally at a faster ratethan surrounding cells, resulting in a preligule band ofsmall cells across the primordium. Preligule band cells

on either side of the midrib then divide periclinally. A‘‘wave’’ of periclinal divisions toward the leaf primor-dium margins and midrib causes an epidermal ridge(ligular ridge) to arise that spans the width of the leafprimordium. The ridge develops into the ligule and cellsdistal to the ridge form the auricles. At maturity, theligule and auricle cells are morphologically distinct fromeach other and from blade and sheath cells.

The transition from sheath to blade involves alterationnot only in the epidermis but also in the pattern of thevasculature (Sharman 1942). Throughout the leaf there isa single midvein with multiple lateral veins on eitherside. Within the blade, several smaller, intermediateveins exist between the laterals and these anastomose(join) at the ligular region. Consequently, the sheath hasonly one intermediate vein between each lateral vein.

Ligule development occurs within the larger contextof leaf primordium development (for review, see Syl-vester et al. 1996). Clonal analysis shows that leaf found-er cells initially divide equally to form the leaf primor-dium (Poethig 1984). These divisions are polar, however,as cells at the future midvein position divide first and

1Corresponding author.E-MAIL [email protected]; FAX (510) 642-4995.

208 GENES & DEVELOPMENT 11:208–218 © 1997 by Cold Spring Harbor Laboratory Press ISSN 0890-9369/98 $5.00; www.genesdev.org

Cold Spring Harbor Laboratory Press on June 2, 2018 - Published by genesdev.cshlp.orgDownloaded from

http://genesdev.cshlp.org/

http://www.cshlpress.com

divisions then follow laterally around the meristem.Such sequential order of development is also observedfor vein development. In the primordium, the first vis-ible vein is the midvein, followed by the sequential ap-pearance of the lateral veins, with the first developingclosest to the midvein. Both the midvein and the lateralveins develop acropetally (base to tip). Intermediateveins branch from the lateral veins and develop ba-sipetally (tip to base). They are visible only after theirbracketing lateral veins have developed entirely to theleaf margins. In the context of vein formation, ligulardivisions are first apparent while intermediate veins aredeveloping basipetally (Sharman 1942; Becraft et al.1990). At this stage, epidermal differentiation is proceed-ing from tip to base but has not yet reached the ligularregion.

Two unlinked recessive null mutants, liguleless1-ref-erence (lg1-R) and liguleless2-reference (lg2-R), alter thedevelopment of the ligular region. lg1-R, located at po-sition 11 on chromosome 2, removes the ligule and au-ricle in all but the uppermost leaves of the plant, whererudimentary ligule forms. Preligule band divisions canoccur on lower leaves, however, they are only in thetransverse anticlinal dimension (Sylvester et al. 1990).The boundary between the sheath and blade is at a po-sition similar to that in wild type but is less distinct(Emerson 1912; Becraft et al. 1990). The lg1-R mutantphenotype is cell autonomous (Becraft et al. 1990) andLG1 may be involved in the reception and/or propaga-tion of a ‘‘make ligule/auricle’’ signal that originates oneither side of the midrib (Becraft and Freeling 1991). LG1shows sequence homology to a novel class of DNA bind-ing proteins (squamosa-promoter binding proteins) andexpresses in the developing ligular region (Moreno et al.1997).

lg2-R, located at position 101 on chromosome 3, has aless severe phenotype than lg1-R (Brink 1933; Harper andFreeling 1996). lg2-R mutants lack ligule and auricleonly on the lower-most leaves of the plant. On higherleaves, ligule and auricle begin to appear at the leaf mar-gins and with each successive leaf the phenotype gradu-ally becomes less severe. Where the ligule and/or auricleis lacking, the sheath can extend into the blade. Mosaicanalysis shows that the lg2-R mutant phenotype is cellnonautonomous (Harper and Freeling 1996), indicatingthe involvement of cell–cell signaling in LG2 function.Double mutant analysis and reciprocal dosage sensitiv-ity studies between Lg1+ and Lg2+ suggest that thesegenes act in a common circuit of action (Harper andFreeling 1996).

Here we describe the early morphological events lead-ing to the development of the lg2-R mutant blade–sheathboundary. SEM analyses indicate that lg2-R mutantleaves do not form a normal preligule band and suggestthat LG2 is functioning either at or before the time ofpreligule band divisions. We also present the cloning andinitial expression studies of Lg2+. We cloned the Lg2+gene by transposon tagging and it shows sequence ho-mology to the basic leucine zipper (bZIP) class of tran-scription factors. It is most similar to a subclass of plant

DNA-binding proteins characterized biochemically andhypothesized to function in diverse processes (Tabata etal. 1991; Kim et al. 1994; Liu and Lam 1994; Qin et al.1994; Ulmasov et al. 1994; Zhang and Singh 1994; Zhanget al. 1995; Chengbin et al. 1996). Lg2+ is the first re-ported gene within this group with an associated mutantphenotype. Reverse transcriptase (RT)–PCR analyses in-dicate that LG2 mRNA accumulation precedes that ofLG1 mRNA.

Results

SEM of developing and mature lg2-R mutant leaves

A SEM study was undertaken to characterize the devel-opment of lg2-R mutant ligular regions as comparedwith wild-type siblings. In this study, we give leaves twodesignations—leaf number (L#) and plastochron number(P#). The leaf number was counted from the base of theplant toward the apex. Leaf number therefore indicatesleaf position on the main axis of the plant. In contrast,plastochron number indicates the relative age of the leaf.The plastochron number was counted from the apicalmeristem toward the base of the plant. The most re-cently initiated leaf primordium was termed P1.

In wild type, no visible sign existed of ligule and au-ricle formation (Fig. 1A) until an increased region of an-ticlinal divisions (Fig. 1B) occurred across the adaxial (in-ner) leaf surface. A subset of these cells then dividedpericlinally (Fig. 1C) to form an outgrowth of epidermalcells that will eventually form the mature ligule. Thesetwo stages of development, the increased anticlinal divi-sions followed by periclinal divisions, have been termedthe preligule and ligule ridge stages (Sylvester et al.1990).

In developing lg2-R mutant leaves, the preligule bandwas often incomplete (Fig. 2A). The number of cellswithin the preligule band decreased from margin to mid-rib and the band ended before the midrib. Similarly, dur-ing the ligule ridge stage the number of periclinal divi-sions decreased along the margin to midrib dimension(Fig. 2C). The ligule ridge decreased in height until itsappearance resembled that of the preligule band. Imme-diately distal, a group of small square cells was observedthat had undergone more anticlinal divisions than sur-rounding cells (Fig. 2C, brackets). This pattern, charac-terized by the ligule diminishing and then displacinginto the blade region, was observed frequently in lg2-Rmutant leaves.

The morphology of mature lg2-R leaves was consistentwith the morphology of its developing leaves. A maturelg2-R mutant leaf 1 is shown in Figure 3A. At the mar-gin, a small group of cells underwent anticlinal as well assome periclinal divisions, but a normal auricle and liguledid not form. This group of cells is likely part of theligule region because of the distal blade and proximalsheath cell morphologies. On the remainder of the leaf,the transition from blade-like to sheath-like cells wasless distinct. The characteristic blade cells apparent atthe top of the figure (Fig. 3A) gradually gave way to the

LG2 functions early during ligule development

GENES & DEVELOPMENT 209




longer cells of the sheath. This is an example of a diffuseblade–sheath boundary.

Figure 3B shows a mature lg2-R mutant leaf 4. A near-normal auricle and ligule were evident at the margin.The ligule, however, gradually shortened from margin tomidrib and finally stopped at the position of a lateralvein. A small group of cells with neither blade nor sheathmorphology was found distally and on the midrib side ofthe vein. This group of cells was composed of small,rounded cells and was pushed up from the surface of theleaf suggesting both anticlinal and periclinal divisionsoccurred in this region. The vein along which the dis-placement occurred separated the shorter, more crenu-

lated blade cells (on the margin side) from the longer,noncrenulated sheath cells (on the midrib side).

Identification and genomic cloning of a lg2transposon-induced allele

A potential Mutator (Mu) transposon-induced allele oflg2, lg2-2757, was identified in a screen of selfed Muactive families (Harper and Freeling 1996). To identifywhich Mu element caused the lg2 phenotype, a cosegre-gation analysis was performed. Genomic DNA from lg2-2757/lg2-R individuals and +/lg2-R siblings was digestedwith a number of different restriction enzymes. South-

Figure 2. lg2-R mutant leaf primordiumligule and auricle development. (A) Leaf 5,plastochron 9: The preligule band (anticli-nal divisions) is incompletely formedacross the primordium width. (B) The tran-sition area boxed in A is displayed athigher magnification. (C) Leaf 4, plas-tochron 10: The ligule ridge is not devel-oped equally across the width of the leafprimordium. A small displacement isshown in brackets. Notice the undulatingsurface reflecting the underlying venationin the blade above the ligule (cf. Fig. 1C)and how this pattern continues above thedisplacement. The arrow under the wordmargin indicates the direction of the leafmargin.

Figure 1. Wild-type ligule and auricle de-velopment. (A) Leaf 7, plastochron 7: An-ticlinal cell divisions giving rise to the lig-ule and auricle have not been initiated. Inthe top third of the image, the outer cellfiles begin to curve toward the margin.The ligule will form below this region.The primordium base can be seen in thelower right-hand corner. (B) Leaf 6, plas-tochron 8: The preligule band (anticlinaldivisions) is formed across the leaf primor-dium. (C) Leaf 5, plastochron 10: Periclinaldivisions are well advanced, causing theligule ridge to protrude from the adaxialleaf surface. Above the ligule ridge, noticethe undulating surface reflecting the un-derlying venation. The arrow under theword margin indicates the direction of theleaf margin.

Walsh et al.

210 GENES & DEVELOPMENT




ern blots were then hybridized sequentially with probesrepresenting the unique interior sequence of the autono-mous element MuDR and the nonautonomous elementsMu1, Mu3, Mu4, Mu5, and Mu8. The Mu8 probe hybrid-ized to a 5.5-kb EcoRI restriction fragment that cosegre-gated with the lg2-2757 allele (data not shown). Wecloned this fragment of DNA because EcoRI does nothave a recognition site within the Mu8 element, therebyyielding unique flanking sequences.

A size-fractionated genomic bacteriophage library wasmade from a lg2-2757/lg2-R mutant individual and∼750,000 clones were screened with the Mu8 probe.Fourteen positive clones were identified and grouped byrestriction mapping. For each class of clones, a proberepresenting non-Mu8 DNA was hybridized to the origi-nal segregating family Southern blots. A clone (Fig 4C;lg2.2757.EE) was identified containing a 1.2-kb EcoRI–BamHI fragment (Fig 4D; lg2.2757.EB) that hybridized tothe same 5.5-kb fragment as the Mu8 element.

Linkage to other lg2 mutant alleles

To insure that the lg2.2757.EE genomic clone repre-

sented part of the lg2 locus and not a linked locus, inde-pendently derived lg2 mutants were analyzed. Five lg2mutations were recovered from Mu-directed tagging ex-periments. Mutant lg2-902 was identified in a screen of10,000 plants (L. Harper and M. Freeling, unpubl.) andlg2-219, lg2-228, lg2-229.1, and lg2-229.2 in a separatescreen of 18,000 plants. All alleles have a phenotypesimilar to lg2-R. After two introgressions into the inbredline W23, lg2-229.1 expressed a dramatically milder phe-notype. Initial characterization of the lg2-229.1 alleleshows a Mu element insertion in the first intron of thegene.

Genomic DNA from the five putative Mu-induced al-leles and their respective progenitor alleles was digestedwith the restriction enzymes EcoRI, BamHI, XhoI, andBglII. Southern blots were hybridized to either thelg2.2757.EB probe or a 1.7-kb BglII–EcoRI probe (Fig. 4D;lg2.2757.BgE). In all cases the new mutant allele is dif-ferent from the wild-type progenitor allele (Fig. 4A). lg2-902, lg2-228, lg2-229.1, and lg2-229.2 carry some form ofrearrangement or insertion and lg2-219 is a deletion. Re-striction mapping places the lg2-902 and lg2-229.1 le-sions 58 and the lg2-229.2 lesion 38 of the lg2-2757 allele

Figure 3. lg2-R mutant plant matureleaves. (A) Leaf 1, plastochron 13: A smallligule/auricle vestige is found at the leafmargin, as indicated by the arrow. Wherethis structure is missing the transitionfrom blade to sheath is diffuse; long, rect-angular sheath cells gradually yield to theshort, square blade cells. (B) Leaf 4, plas-tochron 10: Ligule and auricle are evidentat the leaf margin. These mature structuresare most developed at the margin and be-come decreasingly less developed towardthe midrib. A region of cells is displacedalong a lateral vein, as shown in brackets.The arrow under the word margin indicatesthe direction of the leaf margin.






Mu8 insertion (Fig. 4B). The end points of the lg2-219deletion allele have not been mapped. Neither thelg2.2757.EB nor the lg2.2757.BgE probes, however, hy-bridized to lg2-219 DNA. Both probes hybridize less in-tensely to other restriction fragments—an example is inFigure 4A(c)—indicating Lg2+ is part of a gene family.

Identification of a LG2 cDNA and genomic clones

To identify possible coding sequences, ∼400 bp from the58 and 38 ends of the lg2.2757.EE genomic clone weresequenced. On the basis of a potential 240-bp open read-ing frame (ORF), two primers (lg2.1 and lg2.2) were syn-thesized for a PCR assay (Fig. 4D; lg2 1.2). Using a PCRproduct amplified from the lg2.2757.EE clone as a probe(Fig. 4D; lg2 1.2), ∼1.5 million cDNA clones werescreened and six positive hybridizing clones identified.Sequence at the 38 ends of four of the cDNA clones wasidentical over 360 bp. Three clones revealed two poten-tial polyadenylation sites 39 bp apart (Fig. 5). The longestof these clones (1825 kb) was sequenced in its entirety.The longest ORF predicts a protein of 531 amino acids(Fig. 5). The nucleotide context of the first methionine,

GCCAUGG, is identical to the PuCCAUGG Kozak con-sensus sequence (Kozak 1989). The cDNA contains ashort 58 untranslated sequence of 5 bp and a 38 untrans-lated sequence of 226 bp.

To analyze further the structure of the gene, two lg2genomic BamHI fragments were cloned from the inbredline B73 (Fig. 4C, B73.7 and B73.4). When combined withthe lg2.2757.EE genomic clone, the genomic region ofthe cDNA is almost covered. A 750-bp PCR product (lg242.34) was generated to derive contiguous sequence (Fig.4C). Comparing genomic DNA and cDNA sequencesidentified 11 introns (Fig. 5), all with 58 and 38 consensussplice sites. Intron sizes (Fig. 5) were determined by se-quencing or in the case of introns 1, 4, and 10 by com-parison of genomic PCR product size with the cDNA.

Alignment of the genomic sequence with the 58 end ofthe cDNA sequence allowed the identification of a po-tential promoter region (Fig. 5). Putative TATA (TTT-TATATA) and CAAT boxes (TAACC) are present 49 and89 bp, respectively, upstream of the start of the cDNAsequence (Fig. 5). Four adenines, representing potentialtranscription start sites, are present 14, 20, 27, and 29 bpdownstream of the putative TATA box. These differ

Figure 4. (A) Five putative Mu transposon-in-duced liguleless2 alleles differ from their pro-genitor alleles. (a–d) Restricted DNA (a and c,BamHI; b, XhoI; d, EcoRI) from putative Mu-tagged plants and their respective wild-typesiblings hybridized to the lg2.2757.EB probe. (e)EcoRI-restricted DNA from a lg2-902 homozy-gous plant and plants homozygous for its pro-genitor alleles hybridized to the lg2.2757.EBprobe. The arrows point to the tagged frag-ments. (B) Multiple independently derived al-leles map to the same genomic locus. The po-sitions of lesions are shown on a restrictionmap of the progenitor genomic locus. Knowninsertion positions are indicated by trianglesand the lg2-2757 Mu8 insertion is shaded. Re-striction mapped insertions and/or rearrange-ments are indicated by a triangle beneath a linespanning the possible region. The minimumextent of the deletion allele (lg2-219) is shownby a shaded line. (C) Cloned fragments fromlg2-2757 and B73. The Mu8 position within thelg2-2757 clone is indicated by a triangle (D)DNA fragments used as radiolabeled probes.

Walsh et al.





from the consensus start of transcription, Py2CAPy5, by1, 2, 3, and 3 bp, respectively. The presence of consensus

promoter elements, transcription start sites and a trans-lation initiation codon suggests, but does not prove, thatthe sequence in Figure 5 shows the complete LG2mRNA.

The LG2 cDNA has homology to the bZIP classof transcription factors

A homology search of the GenBank database using theBLAST algorithm (Altschul et al. 1990) identified theLG2 protein as a member of the bZIP class of transcrip-tion factors. The leucine zippers facilitate homo- and/orheterodimerization through hydrophobic interactions,and the adjacent basic domains bind to specific DNAsequences. The LG2 basic domain (residues 224–248) isalmost identical to a plant subgroup of bZIP proteins(Fig. 6A), which is similar to a wide range of eukaryoticbZIP proteins (Fig. 6B,C). Within this subgroup, stronghomology with LG2 continues through the leucine zip-per and ends at the second to last exon of LG2 (residues252–502). No proteins within the GenBank databaseshow homology to the 224 amino-terminal residues andthe 29 carboxy-terminal residues of LG2.

RT–PCR analysis of LG2 and LG1 mRNA in wild-typeand mutant developing ligular regions

To correlate LG2 mRNA expression with ligule and au-ricle development, RT–PCR was used to detect LG2mRNA in developing ligular regions of wild-type, homo-zygous lg1-R (deletion allele), and homozygous lg2-219(deletion allele) plants. LG2 mRNA was detected inwild-type and lg1-R fractions (Fig. 7A). The expression isLG2 specific because no message was detected in thelg2-219 samples.

To determine if LG1 mRNA expression is affected bythe absence of LG2, we performed the complementaryexperiment. LG1 mRNA expression in wild-type and ho-mozygous lg2-219 plants was compared. LG1 expressionis specific because no message was detected in the lg1-Rsamples. LG1 mRNA was detected in wild-type and lg2-219 fractions. In wild type, LG1 mRNA expression isrestricted to leaf primordia undergoing preligule bandand ligule ridge divisions (Fig. 7B). LG1 wild-type expres-sion was assayed previously by RT–PCR (Moreno et al.1997), but the youngest developmental fractions in-cluded leaf primordia undergoing preligule band divi-sions. Therefore, the first developmental time at whichLG1 mRNA is expressed could not be determined.

Discussion

The blade–sheath boundary incompletely formsin lg2-R mutants

The involvement of LG2 in the normal development ofthe blade–sheath boundary has been studied previouslygenetically (Harper and Freeling 1996). To determinewhen LG2 acts in development we examined the devel-oping ligular region of lg2-R mutant plants. We used

Figure 5. The nucleotide and predicted amino acid sequence ofLG2. Numbers indicate amino acid residues. The putativeTATA and CAAT boxes are underscored. The basic region (224–248) is indicated in bold and the three leucines within the leu-cine zipper (248, 255, and 262) are in bold italics. (j) The stopcodon; (.) intron positions. Intron sizes are given as number ofbase pairs. The first and last nucleotides of the sequenced cDNAare in boldface type. Polyadenylation sites are indicated by as-terisks.






SEM analysis to compare the development of lg2-R mu-tant plants with wild-type siblings. The first visible al-teration is the preligule band (anticlinal divisions),which is apparent at the primordium’s margins but doesnot extend across the full width of the primordium. Con-sequently, altered ligule ridge (periclinal divisions) andmature ligule develop. Compared to wild type, thesetruncated structures do not develop equally across theleaf’s width. Additionally, patches of ligule and auriclesometimes displace along lateral veins. All these distur-bances may be explained if LG2 functions at or beforepreligule band divisions. The above lg2-R mutant phe-notypes are the result of perturbations in the initial es-tablishment of the preligule band.

An incorrectly established preligule band could ex-plain the longitudinal displacement, as well as the lat-eral decrease in the extent of ligule/auricle developmentin lg2-R leaves. The phenotypes could be the result ofeither incorrect positioning or a slight delay in normalligule and auricle development, resulting in an uncou-pling from the overall development of the leaf primor-dium. As reviewed in the introduction, several develop-mental events within the leaf primordium occur tempo-rally along both the lateral and longitudinal dimensions.If a leaf primordium’s ligule and auricle development arenot synchronous with other developmental events, thencells may not be equally competent to interpret or carryout determination and differentiation processes. The re-sult is mature structures that do not develop equally andappear at displaced positions. Further investigation isclearly needed to test this hypothesis and to determinethe mechanisms.

lg2-R mutant plants also have a tassel phenotype. Pre-liminary observations show a reduction in lateral tasselbranching and the elongation of the upper vegetative in-ternode can be suppressed. We are able to detect LG2mRNA in immature tassels by RT–PCR. This phenotypesuggests that Lg2+ may be involved both in the laterevents of vegetative development and the early events ofinflorescence development.

lg2 encodes a bZIP transcription factor

We cloned the lg2 locus via Mu transposon tagging andconfirmed its identity by independently derived mutant

alleles. LG2 shows sequence homology to the bZIP tran-scription factor class of proteins. LG2 is most similar toa subclass that has been found in a wide range of plantspecies including Arabidopsis (Kawata et al. 1992; Schin-dler et al. 1992a; Zhang et al. 1993; Miao et al. 1994;Xiang et al. 1995), wheat (Tabata et al. 1991), tobacco(Katagiri et al. 1989), fava bean (Ehrlich et al. 1992), po-tato (Feltcamp et al. 1994), and maize (Foley et al. 1993).Members of this group were first identified by their abil-ity to bind similar promoter sequences, or by having se-quence homology to each other. This bZIP subclassbinds promoter sequences that include the octopine syn-thase (ocs) element within some Agrobacterium promot-ers, the activation sequence (as-1) within the CaMV 35Spromoter and the hexamer (hex) motif of wheat histonegene promoters. A 20-bp DNA-binding consensus se-quence has been identified (Bouchez et al. 1989) that issimilar to the cAMP response element (CRE) (Lin andGreen 1988).

Not only does their binding specificity characterizethis subclass but so does their dimerization specificity.This group of bZIP proteins can form homodimers butcannot heterodimerize with another large subclass ofplant bZIP proteins called the G-box-binding factors (Ta-

Figure 7. RT–PCR analysis of LG2 mRNA and LG1 mRNAabundance in developing ligule regions of wild-type and mutantplants. The plant genotypes are indicated to the left. The lg2-219 and lg1-R alleles are molecular deletions. A and B showPCR fragments of LG2 RT–PCR and LG1 RT–PCR, respectively.The numbers 1, 2, and 3 refer to the RNA samples. (1) P6-8marginal region; (2) P6-8 midrib region; (3) meristem and P1-5;(P) Plastochron.

Figure 6. (A) The LG2 bZIP domain iscompared with closely related plant pro-teins (Katagiri et al. 1989; Tabata et al.1991; Ehrlich et al. 1992; Foley et al. 1993;Feltkamp et al. 1994; Xiang et al. 1995). (B)The LG2 bZIP domain is compared withother bZIP-containing plant proteins(Hartings et al. 1989; Schmidt et al. 1990;Schindler et al. 1992a). (C) The LG2 bZIPdomain is compared with two mammalianbZIP proteins (Hoeffler et al. 1988;Bohman et al. 1987). Identical amino acidsare indicated by dashes, and the leucineresidues within the leucine zipper by as-terisks and boldface type.

Walsh et al.





bata et al. 1991; Schindler et al. 1992). Sequences outsideof the leucine zipper contribute to dimerization specific-ity (Katagiri et al. 1992). Carboxy-terminal sequencesstabilize dimerization and a full-length protein cannotdimerize with a protein lacking these sequences. LG2has homology within the bZIP region as well as the car-boxy-terminal regions of this group of proteins. Regionsof LG2, however, are unique. It is therefore unlikely thatan ortholog has been cloned from another species.

Although many plant bZIP proteins have been identi-fied, few have a corresponding mutant phenotype. Thefirst plant bZIP protein to be assigned a biological func-tion by association with a mutant phenotype was themaize protein OPAQUE2 (Hartings et al. 1989; Schmidtet al. 1990). Possible functions for proteins in the sub-class of bZIPs that LG2 is a member of have been pro-posed by studying the cis-binding sequences either trans-genically, through fusions with reporter genes, or bystudying plant genes with promoters containing the con-sensus sequence. Hypothesized functions include tran-scriptional regulation of plant histone (Tabata et al.1991) and GST genes (Ulmasov et al. 1994; Zhang et al.1995; Chengbin 1996) as well as auxin, salicylic acid, andmethyl jasmonate responses (Kim et al. 1994; Liu andLam 1994; Qin et al. 1994; Zhang and Singh 1994;Chengbin 1996). The Lg2+ gene has the first publishedmutant phenotype associated with this group of proteinsand adds a developmental function to the above list.

Two family members have been cloned previouslyfrom maize, OBF3.1 and OBF3.2 (Foley et al. 1993).These two proteins are closely related and share 95.8%similarity over 338 amino acids. LG2 shares 57% simi-larity to OBF3.2 over 291 amino acids. We have cloned afourth member (J. Walsh and M. Freeling, unpubl.) that ismore similar to LG2 than to either OBF3.1 or OBF3.2. Toobtain mapping information, Foley and coworkers (1993)used the OBF3.1 cDNA as a probe onto recombinant in-bred lines. The probe hybridized with four restrictionfragments, two of which gave a polymorphism. The twofragments mapped to two separate loci, one on chromo-some three at position 105 and another on chromosomeeight at position 075. They were unable to find genespecific probes and therefore were unable to determine ifeither of their clones mapped to these positions. The lg2locus mapped by a three-point cross to chromosome 3 atposition 101 (Brink 1933) and may therefore representthe locus on chromosome 3. The maize chromosomes 3and 8 show colinearity of homeologous sequences (Hel-entjaris et al. 1988) that may be the result of an ancientsegmental allotetraploidy event (Gaut and Doebley1997). Given the mapping data, it is possible that lg2 hasa closely related family member on chromosome 8 andpossibly one on chromosome 3. Whether or not the pu-tative ancient duplication gene is cloned, and whether itis functional, remains to be determined.

The identification of LG2 as a member of a bZIP pro-tein family, raises the possibility that the ligule and au-ricle structures present in lg2 mutants result from re-lated genes either compensating or having a functionalredundancy. Functional redundancy can be of particular

importance in maize because of its ancient duplication(Gaut and Doebley 1997). In maize, two proteins canjointly perform the functions of a single protein in an-other plant species (Mena et al. 1996). Consequently, ifone gene is mutated then the resulting phenotype can beless severe then if an orthologous gene was mutated inanother species. Within animals, compensation has beenseen for a number of transcription factor gene families,including CREB/ATF (Hummler et al. 1994; Blendy et al.1996) and myogenic HLH transcription factors (Rudnickiet al. 1992). If LG2 family members can restore the for-mation of ligule and auricle structures in lg2 mutantsthen this ability must be incomplete or we would notobserve the recessive mutant phenotype. Another possi-bility for the ligule and auricle structures in lg2 mutantplants is transposon mosaicism. This possibility is dis-counted because ligule and auricle are present in plantswith a deletion of the lg2 gene.

LG2 and LG1 may act in a common pathway

RT–PCR analysis indicates that LG2 mRNA accumula-tion precedes that of LG1 mRNA. Harper and Freeling(1996) showed that the lg1-R and the lg2-R mutant phe-notypes are sensitive to the wild-type dose of the recip-rocal gene, suggesting that they act temporally close toeach other. Taken together, these data suggest that LG2functions before LG1. Our data, however, also show thatLG2 mRNA accumulation is neither sufficient nor nec-essary for LG1 accumulation.

The specific role of LG2 in the definition of the blade–sheath boundary and its associated structures is still un-der investigation. We can begin to place its functionwithin development. The cell nonautonomous pheno-type (Harper and Freeling 1996) of LG2 indicates it isfunctioning in a cell–cell signaling pathway. The identi-fication of LG2 as a transcription factor suggests a down-stream signaling function. Recent reports, however,show that some plant transcription factors can movefrom cell to cell (Lucas et al. 1995; Perbal et al. 1996).Whether or not LG2 protein is moving or a downstreammolecule is involved in the signaling process needs to bedetermined. SEM analysis implicated LG1 in early stagesof ligule and auricle determination and mosaic analysisexperiments suggested that it has a role in the propaga-tion or reception of a ‘‘make ligule/auricle’’ signal (Be-craft and Freeling 1991). Our expression data is consis-tent with the idea that LG2 functions before LG1. LG2may induce the make ligule-auricle signal either by fur-ther defining the blade–sheath boundary or by being di-rectly involved in induction of the preligule region.

Materials and methods

Alleles

lg2-R is a spontaneous mutation (Brink 1933) obtained from theMaize Genetics Stock Center. lg2-2757 (also called lg2-rb) wasrecovered in a screen of families from selfed Mu active plants.lg2-219, lg2-228, lg2-229.1, lg2-229.2, and lg2-902 were obtainedfrom two directed Mu tagging experiments. We crossed lg2-R/






lg2-R pollen with +/+ Mu active ears. The progeny was thenscreened for the rare liguleless plant. A mutation in the wild-type progenitor allele, in combination with the mutant refer-ence allele, results in the mutant phenotype. lg1-R is a sponta-neous mutation (Emerson 1912) obtained from the Maize Ge-netics Stock center.

SEM

Dissections were done on 3-week-old seedlings from a popula-tion segregating 1:1 for lg2-R mutant plants and their wild-typesiblings in the Mo17 inbred background. Plants were grown inthe greenhouses of University of California at Berkeley. Dissec-tions were performed under a dissecting microscope. Leaveswere counted in succession, beginning with leaf one and con-tinuing to the apex of the meristem, so that each leaf could beassigned both a leaf and a plastochron number. Each leaf wasremoved by slicing along the midrib and then gently severingthe leaf’s nodal attachment. Leaves were then unrolled onto anadhesive surface with the adaxial surface up. Casts and molds ofthe leaf surfaces were made as described by Sylvester and co-workers (1990). Casts were mounted on aluminum stubs withepoxy glue, coated with 25 nm of gold-palladium with a Polaronsputter coater, and observed using a Topcon ISI DS130 SEMoperating at an accelerating voltage of 10 kV. All images arecomposite images, assembled from two or more 3 × 4-inch SEMPolaroid negatives. Negatives were scanned using a flatbedscanner at 600–800 ppi, then assembled on Adobe Photoshop4.0.

DNA isolation and Southern analysis

Maize genomic DNA was isolated from seedling and adult leaftissues as described (Lisch et al. 1995). Southern analysis wasperformed with the following modifications according toSchneeberger et al. (1995). The hybridization solution contained6× SSC, 2 mM EDTA, 10 mM Tris/HCl (pH 7.5), 5× Denhardts,0.2 mg/ml of salmon sperm DNA, 20 mM sodium phosphatebuffer (pH 7), and 1% N-lauryl-sarkosyl. Blots were washed to afinal stringency of 0.1× SSC and 0.1% SDS at 65°C. A 530-bpDNA fragment internal to the Mu8 terminal inverted repeatswas used for the Mu8 probe. The lg2 probes (lg2.2757.EB andlg2.2757.BgE) are described in Figure 4D.

Genomic cloning

Three genomic libraries were made. For the first library, DNAfrom a lg2-2757/lg2-R mature leaf was digested with EcoRI andenriched by gel electroelution for 5.5-kb fragments. The librarywas constructed in EcoRI predigested l Zap II (Stratagene) vec-tor and packaged using Stratagene’s Gigapack packaging ex-tracts. Nylon or nitrocellulose lifts were hybridized at 65°Cwith the Mu8 probe and washed with 0.1× SSC and 0.1% SDS.Stratagene’s in vivo excision method was used to generate thelg2.2757.EE insert in the pBluescript SK− vector.

Two separate size-fractionated genomic libraries were madefrom the maize inbred B73. Immature ear DNA was digestedwith BamHI and two fractionations of 4 and 7 kb were isolatedby gel electroelution. Both libraries were constructed in theBamHI predigested Lambda Zap Express vector (Stratagene).The 7-kb library was screened with a radio-labeled 58 cDNAprobe (Fig. 5, nucleotide 131–880) and the 4-kb library wasscreened with a 38 cDNA probe (Fig. 5, nucleotide 1538–1956).In vivo excision resulted in the B73.7 insert and B73.4 insert(Fig. 4D) within the pBK–CMV vector.

PCR was used to amplify a 750-bp product from B73 genomic

DNA that contains sequences not included in the above geno-mic clones. The primers used were 58-caaagtcgctaggagcgcag and38-gttttacaccctccctcaccagg. The product was purified by gelelectroelution.

cDNA cloning

As we were not certain of the site of LG2 mRNA expression wechose to make a cDNA library from tissues that included de-veloping ligule and auricle regions. Two-week-old B73 seedlingswere dissected by removing the first two leaves and then cuttingthe seedling below the meristem and above the developing lig-ule of leaf three. The remaining tissues included the meristemand 9–10 whole or partial leaf primordia with their ligular re-gions at different stages of development. Total RNA andPoly(A)+ RNA were isolated according to Kloeckener-Gruissemand coworkers. (1992). The cDNA library was constructed usingthe Lambda Zap II cDNA Synthesis Kit (Stratagene). Primaryclones were screened, as described above for the genomic librar-ies, with the lg2 1.2 probe (Fig. 4D). The lg2 1.2 DNA fragmentwas obtained by PCR (primer 1, gcagtcagtagggagcaccggca;primer 2, gcgggtggtgctgctgaaagctc) using the lg2.2757.EE cloneas template. Clones were in vivo excised into the pBluescriptSK− vector.

RT–PCR

RT–PCR analysis was performed on RNA isolated from 16-day-old plants. We used wild-type and mutant plants homozygousfor either the lg2-219 deletion allele or the lg1-R deletion allele.Leaf primordia (leaf 4–6; plastochron 6–8) were removed fromthe shoot and cut 5 mm above their base. The primordia werethen cut on either side of the thickened midrib region. Themidrib regions of all three plastochrons were pooled, as were themarginal regions. The meristem and remaining five plas-tochrons were used as a third sample. This dissection resultedin three fractions—the margins of primordia undergoing prelig-ule (anticlinal) and ligule ridge (periclinal) divisions, the midribsof the same primordia, and a third fraction that includes themeristem and leaf primordia at stages before the preligule banddivisions. RNA was isolated as described previously (Moreno etal. 1997) for RT–PCR analysis.

RT–PCR was used to detect LG2 and LG1 mRNA. A UBIQ-UITIN RT–PCR reaction was done on all mRNA samples toconfirm their integrity (data not shown). The LG2 and LG1reverse transcription reactions were performed on 2 µg of totalRNA and the UBIQUITIN reactions on 0.5 µg. The LG1 andUBIQUITIN reactions were performed as described previously(Moreno et al. 1997). The same conditions were used for theLG2 reactions. A Lg2 38 untranslated primer (gttttacaccctccct-caccagg) was used in the RT reaction. The PCR reaction wasdone with a 58 primer (tcaacaggccgacaatctaaggc) and a nested 38

primer (gtgaggcctcaaaatccggc) such that the product spans in-tron 11. The RNA product is 248 bp and any contaminatinggenomic DNA is 458 bp. RT–PCR reactions were run on 1.5%agarose gels, Southern blotted, and then hybridized with spe-cific radio-labeled probes.

Sequencing

Plasmid DNA used in sequencing reactions was obtained fromovernight liquid cultures using Qiagen’s QIAprep spin miniprepkit. Sequencing was done at the U.C. Berkeley DNA SequencingFacility with an Applied Biosystems automated sequencer.

Walsh et al.





Acknowledgments

We are especially grateful to members of the Freeling laboratoryfor many helpful discussions and suggestions during the courseof this work. We thank Barbara Kloeckener-Gruissem, LisaHarper, Mark Mooney, Lynne Jesaitis, David Braun, BarbaraLane, and Robin MacDiarmid for critical reading and helpfulcomments on this manuscript. We are grateful to Lisa Harperfor providing genetic strains and Barbara Kloeckener-Gruissemfor providing the mRNA used for the cDNA library. We thankScott Poethig for identifying the lg2 tassel phenotype. We thankthe Berkeley Electron Microscope Laboratory and the CenterFor Biological Imaging for providing facilities and technical ad-vice. This work was supported by a grant from the U.S. Depart-ment of Energy (DOE grant DE-FG03-91ER20028).

The publication costs of this article were defrayed in part bypayment of page charges. This article must therefore be herebymarked ‘‘advertisement’’ in accordance with 18 USC section1734 solely to indicate this fact.

References

Altschul, S.F., W. Gish, W. Miller, E.W. Myers, and D.J. Lipman.1990. Basic local alignment search tool. J. Mol. Biol.215: 403–410.

Becraft, P.W., D.K. Bongard-Pierce, A.W. Sylvester, R.S. Poethig,and M. Freeling. 1990. The liguleless-1 gene acts tissue spe-cifically in maize leaf development. Dev. Biol. 141: 220–232.

Becraft, P.W. and M. Freeling. 1991. Sectors of liguleless-1 tissueinterrupt an inductive signal during maize leaf development.Plant Cell 3: 801–807.

Blendy, J.A., K.H. Kaestner, W. Schmid, P. Gass, and G. Schutz.1996. Targeting of the CREB gene leads to up-regulation of anovel CREB mRNA isoform. EMBO J. 15: 1098–1106.

Bohmann, D., T.J. Bos, A. Admon, T. Nishimura, P.K. Vogt, andR. Tjian. 1987. Human proto-oncogene c-jun encodes a DNAbinding protein with structural and functional properties oftranscription factor AP-1. Science 238: 1386–1392.

Bouchez, D., J.G. Tokuhisa, D.J. Llewellyn, E.S. Dennis, and J.G.Ellis. 1989. The ocs-element is a component of the promot-ers of several T-DNA and plant viral genes. EMBO J. 8: 4197–4204.

Brink, R.A. 1933. Heritable characters in maize XLVI- Liguleless2. J. Heredity 24: 325–326.

Chengbin, X., Z.H. Miao, and E. Lam. 1996. Coordinated acti-vation of as-1-type elements and a tobacco glutathione S-transferase gene by auxins, salicylic acid, methyl-jasmonateand hydrogen peroxide. Plant Mol. Biol. 32: 415–426.

Cocciolone, S.M. and K.C. Cone. 1993. Pl-Bh an anthocyaninregulatory gene of maize that leads to variegated pigmenta-tion. Genetics 135: 575–588.

Ehrlich, K.C., J.W. Cary, and M. Ehrlich. 1992. A broad beancDNA clone encoding a DNA-binding protein resemblingmammalian CREB in its sequence specificity and DNAmethylation sensitivity. Gene 117: 169–178.

Emerson, R.A. 1912. The inheritance of the ligule and auriclesof corn leaves. Nebr. Agric. Exp. Stn. Annu. Rep. 25: 81–88.

Feltkamp, D., R. Masterson, J. Starke, and S. Rosahl. 1994.Analysis of the involvement of ocs-like bZip-binding ele-ments in the differential strength of the bidirectionalmas1‘‘2’’ promoter. Plant Physiol. 105: 259–268.

Foley, R.C., C. Grossman, J.G. Ellis, D.J. Llewellyn, E.S. Dennis,W.J. Peacock, and K.B. Singh. 1993. Isolation of a maize bZIPprotein subfamily: Candidates for the ocs-element transcrip-tion factor. Plant J. 3: 669–679.

Gaut, B.S. and J.F. Doebley. 1997. DNA sequence evidence for

the segmental allotetraploid origin of maize. Proc. Natl.Acad. Sci. 94: 6809–6814.

Harper, L. and M. Freeling. 1996. Interactions of liguleless1 andliguleless2 function during ligule induction in maize. Genet-ics 144: 1871–1882.

Hartings, H., M. Maddaloni, N. Lazzaroni, N. Di Fonzo, M.Motto, F. Salamini, and R. Thompson. 1989. The O2 genewhich regulates zein deposition in maize endosperm en-codes a protein with structural homologies to transcriptionalactivators. EMBO J. 8: 2795–2802.

Helentjaris, T., D. Weber, and S. Wright. 1988. Identification ofthe genomic locations of duplicate nucleotide sequences inmaize by analysis of restriction fragment length polymor-phisms. Genetics 118: 353–364.

Hoeffler, J.P., T.E. Meyer, Y. Yun, J.L. Jameson, and J.F. Ha-bener. 1988. Cyclic AMP-responsive DNA-binding protein:Structure based on a cloned placental complementary DNA.Science 242: 1430–1433.

Hummler, E., T.J. Cole, J.A. Blendy, R. Ganss, A. Aguzzi, W.Schmid, F. Beermann, and G. Schutz. 1994. Targeted muta-tion of the CREB gene: Compensation within the CREB-ATFfamily of transcription factors. Proc. Natl. Acad. Sci.91: 5647–5651.

Katagiri, F., E. Lam, and N.H. Chua. 1989. Two tobacco DNA-binding proteins with homology to the nuclear factor CREB.Nature 340: 727–730.

Katagiri, F., K. Seipel, and N.H. Chua. 1992. Identification of anovel dimer stabilization region in a plant bZIP transcriptionactivator. Mol. Cell. Biol. 12: 4809–4816.

Kawata, T., H. Imada, K. Shiraishi, Y. Okada, Y. Shimura, andM. Iwabuchi. 1992. A cDNA clone encoding hbp-1b homo-log in Arabidopsis thaliana. Nucleic Acids Res. 20: 1141.

Kim, Y., K. Buckley, M.A. Costa, and G. An. 1994. A 20 nucleo-tide upstream element is essential for the nopaline synthase(nos) promoter activity. Plant Mol. Biol. 24: 105–117.

Kloeckener-Gruissem, B., J.M. Vogel, and M. Freeling. 1992.The TATA box promoter region of maize Adh1 affects organ-specific expression. EMBO J. 11: 157–166.

Kozak, M. 1989. Context effects and inefficient initiation atnon-AUG codons in eukaryotic cell-free translation sys-tems. Mol. Cell. Biol. 9: 5073–5080.

Lin, Y.S. and M.R. Green. 1988. Interaction of a common cel-lular transcription factor, ATF, with regulatory elements inboth E1a-inducible and cyclic AMP-inducible promoters.Proc. Natl. Acad. Sci. 85: 3396–3400.

Lisch, D., P. Chomet, and M. Freeling. 1985. Genetic character-ization of the mutator system in maize: Behavior and regu-lation of Mu transposons in a minimal line. Genetics 139:1777–1796.

Liu, X. and E. Lam. 1994. Two binding sites for the plant tran-scription factor ASF-1 can respond to auxin treatments intransgenic tobacco. J. Biol. Chem. 269: 668–675.

Lucas, W.J., S. Bouche-Pillon, D.P. Jackson, L. Nguyen, L. Baker,B. Ding, and S. Hake. 1995. Selective trafficking of KNOT-TED I homeodomain protein and its mRNA through plas-modesmata. Science 270: 1980–1983.

Mena, M., B.A. Ambrose, R.B. Meeley, S.P. Briggs, M.F. Yanof-sky, and R.J. Schmidt. 1996. Diversification of C-functionactivity in maize flower development. Science 274: 1537–1540.

Miao, Z.H., X. Liu, and E. Lam. 1994. TGA3 is a distinct mem-ber of the TGA family of bZIP transcription factors in Ara-bidopsis thaliana. Plant Mol. Biol. 25: 1–11.

Moreno, M.A., L.C. Harper, R.W. Krueger, S.L. Dellaporta, andM. Freeling. 1997. liguleless1 encodes a nuclear-localizedprotein required for induction of ligules and auricles during






maize leaf organogenesis. Genes & Dev. 11: 616–628.Perbal, M.-C., G. Haughn, H. Saedler, and Z. Schwartz-Sommer.

1996. Non-cell-autonomous function of the Antirrinum flo-ral homeotic proteins DEFICIENS and GLOBOSA is exertedby their polar cell-to-cell trafficking. Development122: 3433–3441.

Poethig, R.S. 1984. Cellular parameters of leaf morphogenesis inmaize and tobacco. In Contemporary problems of plantanatomy (ed. R.A. White and W.C. Dickinson), pp. 235–238.Academic Press, New York, NY.

Qin, X.F., L. Holuigue, D.M. Horvath, and N.H. Chua. 1994.Immediate early transcription activation by salicylic acid viathe cauliflower mosaic virus as-1 element. Plant Cell 6: 863–874.

Rudnicki, M.A., T. Braun, S. Hinuma, and R. Jaenisch. 1992.Inactivation of Myod in mice leads to up-regulation of themyogenic HLH gene myf-5 and results in apparently normalmuscle development. Cell 71: 383–390.

Schindler, U., H. Beckmann, and A.R. Cashmore. 1992a. TGA1and G-Box binding factors two distinct classes of Arabidop-sis leucine zipper proteins compete for the G-box-like ele-ment tgacgtgg. Plant Cell 4: 1309–1319.

Schindler, U., A.E. Menkens, H. Beckmann, J.R. Ecker, and A.R.Cashmore. 1992b. Heterodimerization between light-regu-lated and ubiquitously expressed Arabidopsis GBF bZIP pro-teins. EMBO J. 11: 1261–1273.

Schmidt, R.J., F.A. Burr, M.J. Auckerman, B. Burr. 1990. Maizeregulatory gene opaque-2 encodes a protein with a ‘‘leucine-zipper’’ motif that binds to zein DNA. Proc. Natl. Acad. Sci.87: 46–50.

Schneeberger, R.G., P.W. Becraft, S. Hake, and M. Freeling.1995. Ectopic expression of the knox homeo box gene roughsheath1 alters cell fate in the maize leaf. Genes & Dev.9: 2292–2304.

Sharman, B.C. 1942. Developmental anatomy of the shoot ofZea mays L. Ann. Bot. 6: 245–284.

Sylvester, A.W., W.Z. Cande, and M. Freeling. 1990. Divisionand differentiation during normal and liguleless-1 maize leafdevelopment. Development 110: 985–1100.

Sylvester, A.W., L. Smith, and M. Freeling. 1996. Acquisition ofidentity in the developing leaf. Annu. Rev. Cell Dev. Biol.12: 257–304.

Tabata, T., T. Nakayama, K. Mikami, and M. Iwabuchi. 1991.HBP-1a and HBP-1b: Leucine zipper-type transcription fac-tors of wheat. EMBO J. 10: 1459–1468.

Ulmasov, T., G. Hagen, and T. Guilfoyle. 1994. The ocs elementin the soybean GH2-4 promoter is activated by both activeand inactive auxin and salicylic acid analogues. Plant Mol.Biol. 26: 1055–1064.

Xiang, C., Z. Miao, and E. Lam. 1995. Isolation of TGA6 (acces-sion no. L42327), a new member of the TGA family of bZIPtranscription factors in Arabidopsis thaliana (PGR 95-063).Plant Physiol. 109: 721.

Zhang, B. and K.B. Singh. 1994. Ocs element promoter se-quences are activated by auxin and salicylic acid in Arabi-dopsis. Proc. Natl. Acad. Sci 91: 2507–2511.

Zhang, B., R.C. Foley, and K.B. Singh. 1993. Isolation and char-acterization of two related Arabidopsis ocs-element bZIPbinding proteins. Plant J. 4: 711–716.

Zhang, B., W. Chen, R.C. Foley, M. Buttner, and K.B. Singh.1995. Interactions between distinct types of DNA bindingproteins enhance binding to ocs element promoter se-quences. Plant Cell 7: 2241–2252.

Walsh et al.





10.1101/gad.12.2.208Access the most recent version at doi: 12:1998, Genes Dev.

Justine Walsh, Cynthia A. Waters and Michael Freeling

boundary sheath−involved in the establishment of the leaf blade encodes a basic leucine zipper protein liguleless2The maize gene

References

http://genesdev.cshlp.org/content/12/2/208.full.html#ref-list-1

This article cites 47 articles, 25 of which can be accessed free at:

License

ServiceEmail Alerting

click here.right corner of the article or

Receive free email alerts when new articles cite this article - sign up in the box at the top

Cold Spring Harbor Laboratory Press


http://genesdev.cshlp.org/lookup/doi/10.1101/gad.12.2.208

http://genesdev.cshlp.org/content/12/2/208.full.html#ref-list-1

http://genesdev.cshlp.org/cgi/alerts/ctalert?alertType=citedby&addAlert=cited_by&saveAlert=no&cited_by_criteria_resid=protocols;10.1101/gad.12.2.208&return_type=article&return_url=http://genesdev.cshlp.org/content/10.1101/gad.12.2.208.full.pdf

http://genesdev.cshlp.org/cgi/adclick/?ad=41173&adclick=true&url=https%3A%2F%2Fipaper.ipapercms.dk%2FEXIQON%2FMarketing%2Ftechnote%2FNGS-comparison-of-methods-for-microrna-profiling-from-plasma-and-plasma-derived-exosomes%3Futm_source%3DCSHL%26utm_medium



The maize gene liguleless2 encodes a basic leucine...

Documents

Transcript of The maize gene liguleless2 encodes a basic leucine...